Hassle free AWS CodePipeline Slack integration (Terraform module)

This terraform module creates an integration with Slack incoming-webhook to post a message every time there is an update to a CodePipeline pipeline.

CodePipeline events are handled by a lambda function. Code for the lambda function lives on its own github (see https://github.com/felimartina/aws-CodePipeline-slack-integration) and is included in this repo as a submodule so that terraform will clone it along with this module while still keeping and maintaining both, the function code and the terraform module, separated.

Example Messages

IMPORTANT NOTE: Unfortunately this function depends on npm to be installed in the machine doing the provisioning (aka running terraform apply). This is to avoid publishing the node_modules folder to the lambda function repo. If anyone has a better approach I would love to hear it.

So what it basically does is before provisioning the lambda function it will cd into the git submodule holding the function and run npm install -production. After that all it does is just zip the function and provision everything.

I have not tested it yet, but this dependency on npm will likely prevent this module from running on Terraform Enterprise.

Variables (aka params)

• PIPELINE_NAMES: List of CodePipeline pipeline names. This is used for the CloudWatch rule.
• APP_NAME: Give the stack a friendly name. Name will prefix created resources (lambda, role, and policy).
• SLACK_WEBHOOK_URL: Url given by slack when creating an incoming-webhook in your slack account. For details see Slack Documentation.
• SLACK_CHANNEL: Slack channel where you want lambda function to post the message.
• RELEVANT_STAGES: [OPTIONAL] This variable tells the lambda function for which stages it should post a message. The variable must be a comma separated string with different stages. It defaults to SOURCE,BUILD,DEPLOY. This helps you to adjust your integration "verbosity".
• LAMBDA_TIMEOUT: [OPTIONAL] Function timeout in seconds. It defaults to 10 which should be more than enough for this kind of function.
• LAMBDA_MEMORY_SIZE: [OPTIONAL] Function memory size in MB. It defaults to minimum 128 which should be more than enough for this kind of function.

Resources Created

Resources that this module will create are:

• Lambda Function
• latest Alias for lambda function
• IAM Role for lambda function with inline IAM Policy attached
• CloudWatch rule for CodePipeline events
• Integration between CloudWatch and Lambda